Attribution of Conduct in Peace Operations: The ' Ultimate Authority and Control ' Test
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
The European Journal of International Law Vol. 19 no. 3 © EJIL 2008; all rights reserved
..........................................................................................
Attribution of Conduct
in Peace Operations: The
‘Ultimate Authority and
Control’ Test
Kjetil Mujezinović Larsen*
Abstract
The article addresses the issue of whether conduct in international peace operations is attrib-
utable to the troop contributing states or to the United Nations, taking the European Court of
Human Rights’ admissibility decision in the Behrami and Saramati cases as a point of ref-
erence. The Court concluded that conduct by UNMIK and KFOR troops in Kosovo is attrib-
utable to the United Nations. The article examines the content of the ‘ultimate authority
and control’ test that is applied by the Court, and argues that the Court should have taken
a different approach. The Court’s test is in the author’s view difficult to reconcile with the
International Law Commission’s work on the responsibility of international organizations,
with United Nations practice on responsibility for unlawful conduct in peace operations, and
with the Court’s own jurisprudence concerning attribution of conduct to the state. The author
argues further that the Court’s arguments are incomplete even if the Court’s approach were
to be considered correct. The article concludes by expressing concern that the Court’s deci-
sion, when seen in connection with previous case law, in practice renders the European Con-
vention on Human Rights irrelevant in international peace operations.
1 Introduction
When can a state be held accountable for human rights violations committed by mem-
bers of its armed forces during international peace operations? This is the issue that
will be addressed here, taking as a point of reference the European Court of Human
* Research Fellow/PhD Candidate, the Norwegian Centre for Human Rights, the University of Oslo. E-mail:
k.m.larsen@nchr.uio.no
EJIL (2008), Vol. 19 No. 3, 509–531 doi: 10.1093/ejil/chn029510 EJIL 19 (2008), 509–531
Rights’ admissibility decision in the Behrami and Saramati cases,1 which in particu-
lar concerned the issue of attribution of conduct as a requirement for the establish-
ment of accountability. The Court was asked to decide whether actions2 committed by
the NATO Kosovo Force (KFOR) and the United Nations Mission in Kosovo (UNMIK)
constituted violations of the Troop Contributing Nations’ (TCN) obligations under
the European Convention on Human Rights (ECHR). The Court concluded that the
alleged human rights violations were attributable to the United Nations and not to the
individual TCNs, and therefore the Court was not competent ratione personae to exam-
ine the relevant actions. The applications were accordingly declared inadmissible.3
The decision was eagerly anticipated, for several reasons. The case presented the Court
with an opportunity to confirm or depart from the controversial decision in the Banković
case, where an application concerning the NATO bombing of Belgrade in 1999 was also
declared inadmissible.4 It was also the first time that the Court had addressed the issue
of accountability under the ECHR for actions carried out by the armed forces of Con-
tracting States while taking part in a United Nations (UN) mandated peace operation (as
the Banković case, of course, concerned NATO actions in the time prior to authorization
from the UN Security Council). Further, this was the first case before any international
court or tribunal which concerned accountability for human rights violations in a terri-
tory under UN administration. The case also provided the Court with an opportunity to
clarify how it regarded the complex issue of human rights protection in Kosovo. This is
an issue which has caused much concern and debate, also within the scope of the Coun-
cil of Europe, where the Parliamentary Assembly had previously requested an opinion
from the European Commission on Democracy through Law (the Venice Commission).5
The background to the deployment of UNMIK and KFOR in Kosovo is well known,
and need not be elaborated here,6 but the facts of the specific cases should be set out.7
The Behrami case concerned actions by UNMIK and KFOR in the municipality of
Mitrovica in March 2000. While playing, some children found a number of undetonated
1
App. No. 71412/01 Agim Behrami and Bekir Behrami v. France, and App. No. 78166/01 Ruzhdi Saramati
v. France, Germany and Norway, Grand Chamber decision of 2 May 2007 (hereinafter Behrami/Saramati).
2
In this article, the term ‘actions’ also includes omissions, i.e., the failure to act.
3
The decision has been followed up in later cases. In the Court’s decision of 5 July 2007 in App. No.
6974/05 Kasumaj v. Greece, which concerned the occupation by Greek KFOR forces of two plots of land
for the establishment of a main national base, the Court referred to its previous decision, and stated with-
out further discussion that ‘[i]n the light of’ that decision, the application had to be declared inadmis-
sible. See also the decision of 28 Aug. 2007 in App No 31446/02 Gajić v. Germany, which concerned
the refusal of German KFOR troops to pay rent for the use of an apartment that the applicant claimed
ownership of. The application would in any case have been declared inadmissible because the applicant
had failed to exhaust national remedies, but the Court also referred to Behrami/Saramati, supra note 1, as
an independent justification for declaring the application inadmissible.
4
App. No. 52207/99 Banković and others v. Belgium and others, Grand Chamber decision of 12 Dec. 2001.
5
Opinion 280/2004 on Human Rights in Kosovo: Possible Establishment of Review Mechanisms, adopted by
the Venice Commission on 8–9 Oct. 2004, CDL-AD (2004) 033.
6
See in particular UN Security Council res. 1244 (1999). See also Behrami/Saramati, supra note 1, at paras
2–4.
7
The facts of the Behrami case, supra note 1, are described in paras 5–7, and the facts of the Saramati case,
supra note 1, are described in paras 8–17.Attribution of Conduct in Peace Operations: The ‘Ultimate Authority and Control’ Test 511
cluster bomb units, which had been dropped during the NATO bombardment in 1999.
Among the children were two of Agim Behrami’s sons, Gadaf and Bekim. When a
cluster bomb unit exploded, Gadaf was killed while Bekim was seriously injured and
permanently blinded. The application to the Court was founded on Article 2 ECHR,
as it was submitted that French KFOR troops had failed to mark and/or defuse the
undetonated cluster bomb units which the troops knew to be there.8
The Saramati case concerned a Kosovar who was arrested in April 2001 on suspicion
of attempted murder and illegal possession of a weapon. He was released in June, but
he was arrested again in July. His period of detention was repeatedly extended by the
Commander of KFOR (COMKFOR), until he was convicted in January 2002. He based his
application on Article 5 ECHR (both alone and in conjunction with Article 13) and Article
6, as he claimed to have been subject to extrajudicial detention without access to court.9
The application was brought against France and Norway because the COMKFORs who
issued the detention orders were – consecutively – a Norwegian and a French officer.10
In addressing the admissibility issue, the Court took as a starting point that it
was undisputed that Kosovo was not under the control of the Federal Republic of
Yugoslavia at the time of the incidents, and it stated that the territory was ‘under the
effective control of the international presences which exercised the public powers nor-
mally exercised by the Government of the FRY’.11 As such, it appears that the Court
considered that the requirements for extraterritorial application of the ECHR were met.
But the Court continued to say that the relevant question in the case was not primarily
about extraterritorial effect, but whether the Court was competent ratione personae to
examine the states’ contributions to the civil and security presence in Kosovo. In order
to address this issue the Court had to decide whether the conduct could be attributed
to the United Nations.12 In the same manner, this article is therefore not one on the
extraterritorial effect of the ECHR, but rather on the Court’s assessment of attribution
of conduct as an element in establishing accountability for human rights violations. As
will be shown below, however, there are clear connections between these two issues.
The merits of the case will be discussed in Section 2 below, before Section 3 provides
some remarks about whether the decision is a reasonable one, or whether it repre-
sents a set-back to human rights protection in Europe and globally. This section also
briefly addresses a further element in the decision, namely that of the Court’s compe-
tence (or lack thereof) to review conduct which is covered by a UN Security Council
resolution. Finally, in Section 4, some comments are made about the consequences of
the Behrami/Saramati case for the doctrine on the extraterritorial effect of the ECHR.
8
Ibid., at para. 61.
9
Ibid., at para. 62.
10
The application against Germany was based on allegations that a German officer had been involved in Mr
Saramati’s arrest and that Germany was the lead nation in the Multinational Brigade Southeast, where
the detention facilities were located. Based on a lack of evidence concerning the possible involvement of
a German officer, and a recognition by Mr Saramati that German KFOR control over the sector was an
insufficient factual nexus, the case against Germany was withdrawn: see ibid., at paras 64–65.
11
Ibid., at para. 70.
12
Ibid., at para. 71. See also at para. 121.512 EJIL 19 (2008), 509–531
2 To Which Entity is the Conduct of KFOR and UNMIK
Attributable?
A Determining the Proper Test for Attribution
Under Article 2, read together with Article 1, of the Articles on the Responsibility of
States for Internationally Wrongful Acts, adopted by the International Law Commis-
sion in 2001,13 a state may be held responsible for internationally wrongful acts when
an action or omission is attributable to the state and constitutes a breach of an inter-
national obligation of the state. In the Draft Articles on Responsibility of International
Organizations,14 Draft Article 3.2 sets out a corresponding principle for international
organizations. It is the first element in this definition that will be addressed in what fol-
lows, i.e., the attribution of conduct to a state or an international organization.
Chapters II of both ASR and DARIO define in detail in what circumstances con-
duct is attributable to a state or to an international organization, respectively, and
both regimes need to be addressed in order to determine the attribution of conduct
during international peace operations. A natural starting point is Article 4 ASR,
according to which the conduct of a state organ is considered an act of the state
whether the organ exercises legislative, executive, judicial, or any other functions,
whatever position the organ holds within the state and whatever its character is.
A state’s armed forces are an organ of the state, and clearly fall within the scope of
this provision. The situation during peace operations is, however, ordinarily that
military personnel do not act as agents of their home state, but that they are rather
placed at the disposal of the UN or another international organization (e.g., NATO).
This complicates the legal picture.
International responsibility for actions committed during international peace opera-
tions has traditionally been assessed in an ad hoc manner. Ever since the United Nations
operations in Congo (ONUC, 1960–1964) the UN has in practice assumed responsibility
for damage caused by the military forces during peace operations in the performance of
their duties,15 but the legal basis for this practice has largely been unclear. The Behrami/
Saramati case illustrates this point, as the Court does not explicitly make clear what legal
basis it applies in the decision. The Court introduces Draft Article 5 DARIO and Article 6
13
Adopted (with commentaries) by the Commission on its 53rd session in 2001, and endorsed by the UN
GA: see res. 56/83 (28 Jan. 2002) and res. 59/35 (16 Dec. 2004) (hereinafter ‘ASR’). The Articles with
commentaries are also reproduced in J. Crawford, The International Law Commission’s Articles on State
Responsibility (2002).
14
A topic still being considered by the International Law Commission. For reference see: http://untreaty.
un.org/ilc/summaries/9_11.htm (last visited on 20 Feb. 2008). Arts 1–3 with commentaries were pro-
visionally adopted by the Commission during its 55th session in 2003: see Official Records of the General
Assembly, Fifty-eighth Session, Supplement No. 10 (A/58/10), at paras 49–54. Arts 4–6 with commen-
taries were provisionally adopted at the 56th session in 2004: see Official Records of the General Assem-
bly, Fifty-ninth Session, Supplement No. 10 (A/59/10), at paras 64–72 (hereinafter ‘DARIO’).
15
The Special Rapporteur’s Second Report on Responsibility of International Organizations (2004), A/
CN.4/541, at paras 34 ff. It falls well outside the scope of this article to analyse UN practice in this regard.
See K. Schmalenbach, Die Haftung Internationaler Organisationen (2004), at 166–512.Attribution of Conduct in Peace Operations: The ‘Ultimate Authority and Control’ Test 513
ASR (both of which will be addressed below) in the description of ‘relevant law and practice’
(paragraphs 30 and 34), but the Court does not refer to them in its further assessment.
In the legal literature, the prevailing view has been that international responsibility
is linked with operational command over the operations in question. Seyersted argued
that ‘if a Force is under national command, the Organization has no legal responsi-
bility for it and does not represent it internationally’.16 Amrallah stated that ‘the U.N.
would be responsible for the unlawful activities carried out by the armed contingents
put under its disposal by participating states as long as those activities are committed
in the exercise of U.N. functions and under its real and exclusive operational control’
and that ‘[t]he amount of operational control or authority which is exercised over the
U.N. force can be a useful criterion to determine the responsibility of the various par-
ties involved in the peace-keeping operation other than the U.N.’17 Peck has claimed
that ‘[t]he question of who makes the political, strategic, and operational decisions that
together comprise the right to command and control United Nations forces is central
to determining who is responsible for actions taken by U.N. soldiers’.18 Shraga submits
that ‘[i]n enforcement actions carried out by States under the authorization of the Secu-
rity Council … operational command and control is vested in the States conducting the
operation, and so is international responsibility for the conduct of their troops’,19 and
Schmalenbach goes as far as stating that the legal literature is in unison in the view that
‘[d]en Vereinten Nationen ist das Handeln der Peace-keeping-Soldaten zuzurechnen, da sie
während der Einsatzes unter dem operativen Kommando der Vereinten Nationen stehen’.20
The view is also supported by official UN statements, although these indicate that a
distinction must be made according to the legal status of the operations. Peace opera-
tions under the operational control of the UN are regularly given the status of subsidi-
ary organs of the organization,21 and the UN Legal Counsel stated in 2004 that:22
As a subsidiary organ of the United Nations, an act of a peacekeeping force is, in principle,
imputable to the Organization, and if committed in violation of an international obligation
entails the international responsibility of the Organization and its liability in compensation.
16
F. Seyersted, United Nations Forces in the Law of Peace and War (1966), at 411.
17
Amrallah, ‘The International Responsibility of the United Nations for Activities Carried Out by U.N.
Peace-Keeping Forces’ [1976] Revue Egyptienne de Droit International 57, at 65–66.
18
Peck, ‘The U.N. and the Laws of War: How can the World’s Peacekeepers Be Held Accountable?’, 21
Syracuse J Int’l L & Com. (1995) 283, at 293.
19
Shraga, ‘The United Nations as an Actor Bound by International Humanitarian Law’, in L. Condorelli
et al. (eds), The United Nations and International Humanitarian Law (1996), at 330.
20
Schmalenbach, supra note 15, at 249 (original footnotes omitted): ’[t]he conduct of peacekeeping soldiers
is attributable to the United Nations when they are under the operational command of the United Na-
tions’ (author’s translation).
21
Comments and observations received from international organizations (2004), A/CN.4/545, at 18:
‘[t]he principle of attribution of the conduct of a peacekeeping force to the United Nations is premised on
the assumption that the operation in question is conducted under United Nations command and control,
and thus has the legal status of a United Nations subsidiary organ’.
22
Unpublished letter of 3 Feb. 2004 from the UN Legal Counsel to the Director of the UN Office of Legal Af-
fairs’ Codification Division, quoted in A/59/10, supra note 14, at 112, para 5.514 EJIL 19 (2008), 509–531
In 1996 the Secretary-General stated that the international responsibility of the UN
for combat-related activities of UN forces ‘is premised on the assumption that the oper-
ation in question is under the exclusive command and control of the United Nations’,
and that ‘[w]here a Chapter VII-authorized operation is conducted under national
command and control, international responsibility for the activities of the force is
vested in the State or States conducting the operation’.23
In the International Law Commission’s work on the responsibility of international
organizations, the attribution of conduct in peace operations has received consider-
able attention. While Draft Article 4 DARIO establishes the same principle, mutatis
mutandis, for international organizations as Article 4 ASR does for states,24 it is Draft
Article 5 DARIO that governs situations where an organ of a state is placed at the
disposal of an organization. The provision reads:
The conduct of an organ of a State or an organ or agent of an international organization that
is placed at the disposal of another international organization shall be considered under inter-
national law an act of the latter organization if the organization exercises effective control over
that conduct.
The corresponding provision in the ASR is Article 6, which provides that the conduct
of an organ placed at the disposal of a state by another state shall be considered an act
of the former state if the organ is acting in the exercise of elements of the governmental
authority of that state. It should be noted here that the reference in Article 6 ASR to
‘the governmental authority of the State’ necessitated an amendment with regard to
the responsibility of international organizations, as these do not possess any govern-
mental authority.25 The ILC has instead opted for a reference to the ‘effective control’
over the conduct. This is a similar test to that used in Article 8 ASR, which states that
the conduct of a person (or entity) is considered an act of the state if the person ‘is in
fact acting on the instructions of, or under the direction or control of, that State’.26
This requirement of ‘direction or control’ in Article 8 ASR points, in particular, to
the ‘effective control’ test which was applied by the International Court of Justice in
the Nicaragua27 and Genocide cases,28 and, to a lesser extent, the ‘overall control’ test
which was applied by the Appeals Chamber in the International Criminal Tribunal for
the Former Yugoslavia in the Tadić case.29 Following the ICJ’s judgment in the Geno-
cide case, it appears that the application of the ‘overall control’ test is restricted only to
23
Report of the Secretary-General (1996), A/51/389, at para. 17.
24
Draft Art. 4.1 reads as follows: ‘[t]he conduct of an organ or agent of an international organization in the
performance of functions of that organ or agent shall be considered as an act of that organization under
international law whatever position the organ or agent holds in respect of the organization.’
25
A/59/10, supra note 14, at 111 para. 3.
26
DARIO does not include a similar provision to Art. 8 ASR.
27
Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States of America), merits,
judgment of 27 June 1986 [1986] ICJ Rep 14.
28
Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herce-
govina v. Serbia and Montenegro), judgment of 26 Feb. 2007, not yet reported.
29
ICTY Appeals Chamber judgment of 15 July 1999, Prosecutor v. Duško Tadić. See Crawford supra note 13,
at 110–111.Attribution of Conduct in Peace Operations: The ‘Ultimate Authority and Control’ Test 515
the use to which it was originally put in the Tadić case, namely that an armed conflict
may be qualified as international if a state exercises overall control over a group that
is involved in an otherwise non-international armed conflict on another state’s terri-
tory. The test cannot, as it has been speculated,30 be used to establish state responsibil-
ity.31 In the ICJ’s view, therefore, the ‘effective control’ test in Nicaragua is the proper
test under Article 8 of the Articles on State Responsibility.32
In the commentary on Draft Article 5 DARIO, the ILC points out that the control
criterion plays a different role in the context of the responsibility of international
organizations from the one it plays in the context of state responsibility, because it
does not concern the issue whether certain conduct is attributable at all to a state or
an international organization, but rather to which entity – the state or the interna-
tional organization – the conduct is attributable.33 It is therefore not clear whether
the ‘effective control’ test in Draft Article 5 DARIO is to be interpreted identically to
the corresponding test in Article 8 ASR. In the commentary on the provision – and as
cited by the Court – the test is described as follows:34
(1) Article 5 deals with the different situation in which the lent organ or agent still acts to a
certain extent as organ of the lending State or as organ or agent of the lending organization.
This occurs for instance in the case of military contingents that a State placed at the disposal
of the United Nations for a peacekeeping operation, since the State retains disciplinary powers
and criminal jurisdiction over the members of the national contingent. In this situation the
problem arises whether a specific conduct of the lent organ or agent has to be attributed to the
receiving organization or to the lending State or organization.
(6) Practice relating to peacekeeping forces is particularly significant in the present context
because of the control that the contributing State retains over disciplinary matters and criminal
affairs. This may have consequences with regard to attribution of conduct. … Attribution of con-
duct to the contributing State is clearly linked with the retention of some powers by that State over
its national contingent and thus on the control that the State possesses in the relevant respect.
(7) As has been held by several scholars, when an organ or agent is placed at the disposal of an
international organization, the decisive question in relation to attribution of a given conduct
appears to be who has effective control over the conduct in question.
In describing the ‘effective control’ test under Draft Article 5 DARIO as a question of
effective control over specific conduct, it appears that the ILC intends the test to be
similar to the test under Article 8 ASR. The organization must exercise effective con-
trol over the conduct of an organ of a state that is placed at the organization’s disposal
in order for the conduct to be attributable to the organization.
30
Sassòli, ’State Responsibility for Violations of International Humanitarian Law’, 84 IRRC (2002) 401, at
408.
31
Genocide case, supra note 28, at paras 402–407.
32
See, however, Cassese, ‘The Nicaragua and Tadić Tests Revisited in Light of the ICJ Judgment on Geno-
cide in Bosnia’, 18 EJIL (2007) 649, at 665, who argues that the ICJ failed to pay sufficient attention to
state practice and case law.
33
A/59/10, supra note 14, at 111, para. 4.
34
The Behrami/Saramati case, supra note 1, at paras 31 ff; A/59/10, supra note 14, at 110–115.516 EJIL 19 (2008), 509–531
For the sake of completeness, it might also be interesting to note what the Court does
not cite from the commentary. Paragraph (3) of the commentary reads as follows:
The criterion for attribution of conduct either to the contributing State or organization or to
the receiving organization is based according to article 5 on the factual control that is exercised
over the specific conduct taken by the organ or agent placed at the receiving organization’s
disposal. Article 6 of the draft articles on Responsibility of States for internationally wrongful
acts takes a similar approach, although it is differently worded. According to the latter article,
what is relevant is that ‘the organ is acting in the exercise of elements of the governmental
authority of the State at whose disposal it is placed’. However, the commentary to article 6 of
the draft articles on Responsibility of States for internationally wrongful acts explains that, for
conduct to be attributed to the receiving State, it must be ‘under its exclusive direction and con-
trol, rather than on instructions from the sending State’. At any event, the wording of article
6 cannot be replicated here, because the reference to ‘the exercise of elements of governmental
authority’ is unsuitable to international organizations.
Two elements appear from this. First, attribution must be based on factual control,
and be assessed with regard to the specific conduct in question. This is a common
requirement in the ‘effectve control’ test. And, secondly, the commentary’s reference
to Article 6 ASR may indicate that the ‘exclusive direction and control’ criterion also
applies, i.e., that the conduct is attributed to the organization only if the organization
exercises this exclusive direction and control. This criterion of ‘exclusive control’ is not
the same as the ‘effective control’ test, as the former concerns overall control over
an operation rather than control over specific conduct. It appears from the text that
a notion of exclusive control is also included in Draft Article 5 DARIO. This may be
supported by paragraph (5) of the commentary, which points out that ‘[t]he United
Nations assumes that in principle it has exclusive control over the deployment of
national contingents in a peacekeeping force’.
The commentary states further (in paragraph (8)) that:
What has been held with regard to joint operations … should also apply to peacekeeping opera-
tions, insofar as it is possible to distinguish in their regard areas of effective control respectively
pertaining to the United Nations and the contributing State. While it is understandable that,
for the sake of efficiency of military operations, the United Nations insists on claiming exclu-
sive command and control over peacekeeping forces, attribution of conduct should also in this
regard be based on a factual criterion.
It seems therefore to be accepted under these principles that even if the UN claims
exclusive command and control over the peacekeeping forces, specific conduct may
still be attributable to the TCN if the state has effective control over that conduct. Here
the commentary provides a relevant example from the UNOSOM II operation in Soma-
lia from 1993 to 1995.35 The Report of the Commission of Inquiry states that:
The Force Commander of UNOSOM II was not in effective control of several national con-
tingents which, in varying degrees, persisted in seeking orders from their home authorities
before executing orders of the Forces Command. Many major operations undertaken under
the United Nations flag and in the context of UNOSOM’s mandate were totally outside the com-
mand and control of the United Nations …
35
A/59/10, supra note 14, at para. 7.Attribution of Conduct in Peace Operations: The ‘Ultimate Authority and Control’ Test 517
The commentary states that this conduct ‘would be difficult to attribute to the United
Nations’.
A summary is required. It is submitted that if one applies Article 8 ASR in order
to assess the responsibility of the TCNs, the proper test is whether the state exercised
‘effective control’. It is further submitted that if one applies Draft Article 5 DARIO in
order to assess the responsibility of the UN, the proper test is still whether the UN exer-
cised ‘effective control’ over the specific conduct in question, which must be assessed
on the basis of factual criteria. It also appears that even if attribution is assessed in the
light of Draft Article 5 DARIO, one must add the further element that attribution to
the state may be established if the state retains certain powers over its forces.
This brings me to the issue of possible dual or multiple attribution. The question is
whether any given conduct must be attributed to one entity only, or whether it may
be attributed to two or more entities, e.g., the UN, NATO, and/or one or more TCNs
simultaneously. This issue is addressed in the commentaries on ASR as well as on
DARIO. In the commentary on Article 6 ASR, it is clearly stated – albeit briefly – that
dual attribution may occur:36
Situations can also arise where the organ of one State acts on the joint instructions of its own
and another State … In these cases, the conduct in question is attributable to both States under
other articles of this Chapter.
The same point is explicitly made in the commentary on DARIO:37
Although it may not frequently occur in practice, dual or even multiple attribution of con-
duct cannot be excluded. Thus, attribution of a certain conduct to an international organiza-
tion does not imply that the same conduct cannot be attributed to a State, nor does vice versa
attribution of conduct to a State rule out attribution of the same conduct to an international
organization.
One can thus argue that even if given conduct is attributable to the UN, this does not
in itself rule out attribution also to NATO or to one or more TCNs. This aspect is, how-
ever, missing in the Court’s assessment.
Before we return to the case at hand, two further remarks need to be made: First, does
DARIO and/or ASR provide the proper legal basis when the European Court of Human
Rights assesses attribution? And, secondly, are there any specific circumstances that
warrant another threshold for attribution of conduct under human rights law? I will
address these issues only briefly.
The first issue concerns the possible status of ASR and DARIO as customary inter-
national law. A good case can be made that the ASR are in whole or at least in part
an expression of international customary law. In the Genocide case, the ICJ explic-
itly refrains from addressing the customary law status of the ASR as such, as it was
considered unnecessary in that case.38 The ICJ nevertheless indicates that Articles 4
and 8 ASR reflect international customary law, and it has previously indicated in its
36
Crawford, supra note 13, at 103.
37
A/59/10, supra note 14, at 101.
38
Supra note 28, at para. 414.518 EJIL 19 (2008), 509–531
Advisory Opinion in the Cumaraswamy case that Article 6 ASR also reflects custom-
ary law.39 But more relevant at present is that it is doubtful whether Draft Article 5
DARIO can be said to reflect international customary law. While certain elements of
the responsibility of international organizations may have acquired this status, there
seems not to be sufficient practice either by states or by international organizations
with regard to Draft Article 5 DARIO to enable one to draw this conclusion. Most of
the practice concerning attribution of conduct of state organs being placed at the
disposal of international organizations concerns peacekeeping forces,40 and it is at
least difficult to claim the status of international customary law for a rule that is not
supported by this practice. As has been shown above, the adoption of the ‘effective
control’ test seems not wholly to reflect the complexities of the existing practice.
Turning to the second issue, it is unresolved whether the general rules on inter-
national responsibility apply to human rights treaties. Article 55 ASR provides a lex
specialis rule, according to which the rules do not apply ‘where and to the extent that
the conditions for the existence of an internationally wrongful act or the content of the
implementation of the international responsibility of a State are governed by special
rules of international law’. Regardless of whether one considers that human rights
treaties in their entirety are excluded from the scope of application of the general rules
on international responsibility,41 or that these general rules are applicable insofar as
the secondary rules under the human rights treaties are non-existent or ineffective,42
it should be examined whether the human rights instruments provide independent
rules and principles concerning attribution of conduct.
The various human rights instruments require – albeit implicitly – that an action
be attributable to the state for responsibility to be established: see, for instance, Arti-
cle 1 ECHR which places the duty to respect and secure human rights on the states
(‘[t]he High Contracting Parties shall secure to everyone within their jurisdiction …’).
It has been argued that the supervisory organs under the human rights conventions
increasingly find degrees of state involvement not rising to the level established under
the ASR sufficient to render the state responsible.43 The Loizidou case provides a useful
39
Difference Relating to Immunity from Legal Process of a Special Rapporteur of the Commission on Human Rights
(Advisory Opinion), 29 Apr. 1999 [1999] ICJ Rep 62, at para. 62.
40
The Special Rapporteur’s Second Report on Responsibility of International Organizations, 2 Apr. 2004 (A/
CN.4/541), at paras 34 ff. UN practice is described in Schmalenbach, supra note 15; see also, inter alia,
Shraga, ‘UN Peacekeeping Operations: Applicability of International Humanitarian Law and Responsi-
bility for Operations-Related Damage’, 94 AJIL (2000) 406, at 406–412, C.F. Amerasinghe, Principles of
the Institutional Law of International Organizations (2nd edn, 2005), at ch. 12; M. Zwanenburg, Account-
ability of Peace Support Operations (2005).
41
A. Clapham, Human Rights Obligations of Non-State Actors (2006), at 318.
42
Simma and Pulkowski, ‘Of Planets and the Universe: Self-contained Regimes in International Law’,
17 EJIL (2006) 483, at 485; Simma, ‘Human Rights and State Responsibility’, in A. Reinisch and U.
Kriebaum (eds), The Law of International Relations (2007), at 365.
43
Cerone, ‘Human Dignity in the Line of Fire: The Application of International Human Rights Law During
Armed Conflict, Occupation, and Peace Operations’, 39 Vanderbilt J Transnat’l L (2006) 1, at 13 (also
published, with minor amendments, as Cerone, ‘Human Rights on the Battlefield’, in S. Lagoutte et al.
(eds), Human Rights in Turmoil: Facing Threats, Consolidating Achievements (2007), at 97–131).Attribution of Conduct in Peace Operations: The ‘Ultimate Authority and Control’ Test 519
illustration.44 Here the Court found that Turkey exercised ‘effective overall control’
over Northern Cyprus, and that it was therefore unnecessary to determine whether
‘Turkey actually exercises detailed control over the policies and actions of the authori-
ties’ in the territory,45 i.e., that a state’s jurisdiction over a territory in itself means
that the actions of the authorities in that territory are attributable to the state referred
to.46 This ‘merger’ between the ‘effective control’ test and the ‘overall control’ test into
an ‘effective overall control’ test provides a test for attribution that is different from
the ‘effective control’ test in itself. As mentioned, the ‘effective control’ test refers to
specific conduct, while the Court’s test in Loizidou was rather based on overall control
over a territory, which – as also mentioned – following the ICJ’s judgment in the Geno-
cide case does not seem to be a proper test for attribution of conduct under the general
rules on international responsibility.
Other examples of a lower threshold concern the states’ so-called positive (or affirm-
ative) obligations, i.e., that states may be held responsible under human rights law for
private persons’ actions if the states have not taken sufficient measures to prevent the
given action. This doctrine is well acknowledged in the Court’s case law,47 as well as
in the practice of other international human rights tribunals. For instance, the Inter-
American Court of Human Rights stated that the ‘sole requirement is to demonstrate
that the State authorities supported or tolerated infringement of the rights recognized in
the Convention’,48 and, equally relevantly, that:
An illegal act which violates human rights and which is initially not directly imputable to a
State (for example, because it is the act of a private person or because the person responsible
has not been identified) can lead to international responsibility of the State, not because of
the act itself, but because of the lack of due diligence to prevent the violation or to respond to it as
required by the Convention.49
I do not claim on this basis that a state can be held responsible under human rights
instruments for having ‘supported or tolerated’ human rights violations by the UN,
nor do I say that actions by the UN should be assessed like actions by private persons.
A further analysis of these issues falls outside of the scope of this article, but it should
be recalled that, while the question with regard to acts by private persons is whether
the acts are attributable at all to the state, the question when it comes to acts during
44
App. No. 15318/89 Loizidou v. Turkey, judgment (merits), 18 Dec. 1996.
45
Ibid., at para. 56.
46
Ibid., at para 57: ‘[i]t follows from the above considerations that the continuous denial of the applicant’s ac-
cess to her property in Northern Cyprus and the ensuing loss of all control over the property is a matter which
falls within Turkey’s “jurisdiction” within the meaning of Article 1 and is thus imputable to Turkey.’
47
For a brief description see, e.g., C. Ovey and R.C.A. White, The European Convention on Human Rights (4th
edn, 2006), at 28–32 and 51–52 with further references, or (in Norwegian) A. Mirmotahari, EMK og
mellomprivate konflikter (2007), in particular at 38–45. For a general presentation not restricted to the
ECHR see Chirwa, ‘The Doctrine of State Responsibility as a Potential Means of Holding Private Actors
Accountable for Human Rights’, 5 Melbourne J Int’l L (2004) 1.
48
Paniagua-Morales et Al. v. Guatemala, judgment of 8 Mar. 1998, at para. 91 (emphasis added). See also the
Inter-American Commission on Human Rights’ report of 6 Apr. 2001 in Case 62/01 Riofrío massacre, at
para. 48.
49
Velásquez-Rodríguez v. Honduras, judgment of 29 July 1988, at para. 172 (emphasis added).520 EJIL 19 (2008), 509–531
peace operations is whether the acts are attributable to the state or to the UN. But
although this is a relevant distinction, it does not without further clarification from
the Court explain satisfactorily why the principle of holding a state responsible for
the failure to prevent a human rights violation should not apply at all, insofar as the
state is in fact in a position to ‘prevent the violation or to respond to it’. This is relevant
where the UN does not exercise operational command and control, with KFOR as a
clear example. When a human rights infringement occurs through KFOR actions,
the Member States of NATO are undoubtedly in a position to prevent the violation
or to respond to it, either through national orders – where the state has retained this
authority – or through their involvement in NATO itself.
To conclude, this author submits that the correct approach in Behrami/Saramati
would have been (i) to apply the ‘effective control’ test, regardless of whether one
regards the proper question to be whether the troops were placed at the disposal of
the UN or whether they were acting under the direction or control of the UN or the
TCNs, (ii) if it is concluded that the conduct is attributable to the UN, then to consider
whether the situation calls for dual or multiple attribution, and (iii) that in assessing
dual or multiple attribution, to apply the seemingly lower threshold for attribution of
conduct under human rights law.
B An Analysis of the Court’s Approach
This was, however, not the Court’s approach. The Court initially concluded that the
issuing of detention orders fell within the mandate of KFOR and that the supervision of
de-mining fell within the mandate of UNMIK,50 and that the mandates were properly
based on a resolution under Chapter VII of the UN Charter.51 Considering that UNMIK
‘was a subsidiary organ of the UN created under Chapter VII’, the Court concluded that
the actions of UNMIK were attributable to the UN.52 This is not controversial, as the UN
itself assumes responsibility for conduct during operations with the status of subsidiary
organs.53 The Court’s reasoning about KFOR is, however, more controversial.
The key question put forward by the Court was whether the Security Council
‘retained ultimate authority and control so that operational command only was
delegated’.54 The Court’s interpretation of SC Resolution 1244 (1999) was that the
Security Council retains ultimate authority and control over KFOR, and that opera-
tional command was delegated to NATO.55 While it was acknowledged that the TCNs
50
Supra note 1, at para. 127.
51
Ibid., at para. 130.
52
Ibid., at para. 143.
53
Note, however, that the UN in its observations to the Court concluded that the actions in the cases could
not be attributed to UNMIK: see ibid., at para. 120. This is a different issue, as it concerns the relationship
between KFOR and UNMIK, not between the states and the UN. The Court concluded (at para. 127) that
the supervision of de-mining fell within UNMIK’s mandate, and seemingly therefore that the actions in
the Behrami case were attributable to UNMIK. This is not the central issue in this article.
54
Supra note 1, at para. 133.
55
Ibid., at para. 135.Attribution of Conduct in Peace Operations: The ‘Ultimate Authority and Control’ Test 521
retained some authority over their troops, the Court found the essential question to
be whether NATO’s operational command was ‘effective’.56 Answering this question
in the affirmative, the Court concluded that ‘KFOR was exercising lawfully delegated
powers of the UNSC so that the impugned action was, in principle, “attributable” to
the UN’.57
At least three questions arise out of this. One, what is the relationship between the
‘ultimate authority and control’ test that is applied by the Court, and the ‘effective
control’ (alternatively ‘exclusive control’) test? Two, taking into consideration the
similarities between the ‘occupation’ in the Loizidou case and the ‘administration’ in
Kosovo, why does the Court not discuss the ‘effective overall control’ test? And, three,
why does the Court not address the possibility of dual attribution? These questions will
be addressed in this order below.
The Court does not explicitly provide the legal basis for its ‘ultimate authority and
control’ test. This expression does not occur in any case law from the Court itself;58 it
is not used in the commentaries on ASR or DARIO, nor is it used by the ICJ in its case
law. The Court adopted instead an approach that has been advocated in legal litera-
ture concerning the legality of delegations from the Security Council.59 The argument
is that in order for a delegation to be lawful, the Council must at all times retain overall
authority and control over the exercise of the delegated powers,60 and that interna-
tional responsibility for the exercise of powers cannot be transferred as this rests with
the entity to which powers were initially given.61 Sarooshi submits that ‘the ques-
tion of who exercizes operational command and control over the force is immaterial
to the question of responsibility. The more important enquiry is who exercizes overall
authority and control over the forces.’62 He continues by stating that ‘acts of forces
authorized by the Council are attributable to the UN, since the forces are acting under
UN authority’, and that the only two exceptions to this principle are cases where the
Council is ‘prevented from exercizing overall authority and control over the force’, or
when the forces act ultra vires.63
The Court thus has theoretical support for its approach, but it is surprising that the
Court does not attach any comments either on the large amount of literature which
links attribution of conduct with operational control, or on the ‘effective control’ test
in Draft Article 5 DARIO, or on the official UN statements concerning UN practice.
56
Ibid., at para. 138.
57
Ibid., at paras 140–141.
58
However, in App. No. 24833/94 Matthews v. United Kingdom, judgment of 18 Feb. 1999, the UK Par-
liament’s ‘ultimate authority’ to legislate for Gibraltar may have been a relevant factor when the Court
concluded that the UK was responsible for securing the rights guaranteed by Art. 3 of Prot. 1 to the ECHR
in Gibraltar (see at para. 8).
59
See the references in ibid., para. 130.
60
D. Sarooshi, The United Nations and the Development of Collective Security (1999), at 34. See also E. de Wet,
The Chapter VII Powers of the United Nations Security Council (2004), at 265–268.
61
Sarooshi, supra note 60, at 23, n. 89.
62
Ibid., at 163 (original emphasis).
63
Ibid., at 165. See also at 250, n. 8, where the author states that the responsibility for acts delegated to
regional arrangements (e.g., NATO) must be assessed in the same manner as acts delegated to states.522 EJIL 19 (2008), 509–531
It is evident that the ‘ultimate authority and control’ test bears no resemblance to
the ‘effective control’ test. It is not linked with any direct control over a specific action,
and it is not linked with operational command and control. This follows from the
Court’s description of the test: whether the UN Security Council had retained ultimate
authority and control ‘so that operational command only was delegated’. It is further
evident that the test does not require exclusive control, which the Security Council
clearly did not have over KFOR, and it appears that the test is also different from an
‘overall control’ test. The Court does not explain why it addresses ‘ultimate’ control
rather than ‘overall’ control, which is a more common test in international case law
as well as in the legal literature to which the Court referred.
As mentioned above, the Court developed an ‘effective overall control’ test in the
Loizidou case in order to conclude that actions carried out in the occupied territory
in Northern Cyprus were attributable to Turkey. An implicit premise here is that the
occupying power controls the authorities in the occupied territory. Although the
term ‘occupation’ is consistently avoided by the UN when describing the situation
in Kosovo, the UN administration bears many similarities to an occupation, in that
an outside power carries out all functions of government in the territory. Shraga has
stated that ‘in the administration of Kosovo … principles analogous to those of the
laws of occupation applied to questions of respect for the local law and international
human rights standards’.64 But the Court does not discuss whether an analogy could
be drawn with regard to attribution of conduct. The differences between an occupa-
tion and the UN transitional administration do not in themselves explain why the very
test of attribution, i.e., ‘effective overall control’, could not be used similarly. It appears
that the Court considered this test to be inappropriate in the case, and accordingly it
must be assumed that ‘ultimate authority and control’ must have different content
from the test in the Loizidou case.65 If the Court had argued that the UN maintained
effective overall control over Kosovo, then there would at least have been consistency
in the case law. Now it appears that effective overall control is the correct test if a
Contracting State occupies a foreign territory, but not if more Contracting States have
been authorized by the UN to administrate a foreign territory.
Let me then turn to the five factors put forward by the Court in support of its conclu-
sion that the UN retains ultimate authority and control:66 one, that the UN Charter
Chapter VII allows the Security Council to delegate certain powers to other entities;
two, that the relevant power was a delegable power; three, that the delegation was
64
Shraga, ’Military Occupation and UN Transitional Administrations – the Analogy and its Limitation’,
in M.G. Kohen (ed.), Promoting Justice, Human Rights and Conflict Resolution through International Law
(2007), 479, at 487.
65
This point is further confused by the Court’s decision of 16 Oct. 2007 in App. No. 36257/04 and several
others, Berić and others v. Bosnia-Herzegovina, which concerned acts by the High Representative in Bos-
nia-Herzegovina. The question here was ‘whether the UNSC, in delegating its powers … retained effective
overall control’ (emphasis added). In presenting this test the Court referred to Behrami/Saramati, supra
note 1 (i.e. the ‘ultimate authority and control’ test) and Draft Art. 5 DARIO (i.e., the ‘effective control’
test), but not to the Loizidou case, supra note 44.
66
Behrami, supra note 1, at para. 134.Attribution of Conduct in Peace Operations: The ‘Ultimate Authority and Control’ Test 523
explicit in the resolution; four, that the resolution established sufficiently defined limits
on the delegation; and, five, that the leadership of KFOR was required to report to the
Security Council.67 These criteria by no means amount to ‘effective control’ or ‘exclu-
sive control’, and as a test of attribution they seem to be very wide. The only explicit
control mechanism is that KFOR should report to the Security Council. The generaliza-
tion is that if the delegation from the Security Council is lawful and sufficiently defined,
then all actions carried out in accordance with the delegation are attributable to the
UN as long as the delegate has an obligation to report to the Security Council.
It seems to be of significance for the Court that the Security Council has the power
to revoke the delegation.68 As the Court points out, SC Resolution 1244 is drafted
differently in this regard from other resolutions concerning the establishment of
peacekeeping forces. While many operations are given fixed time limits that must be
actively renewed, SC Resolution 1244 determines that the operation shall continue
until it is actively revoked. With the permanent members’ veto power in mind, it is
evident that the Security Council’s control is actually less over KFOR than over other
peacekeeping forces.
The Court also states that direct operational command from the Security Coun-
cil is not a requirement of Chapter VII resolutions. As the Court has concluded that
operational command is not decisive for the attribution of conduct, this becomes
nothing more than an element in the discussion of whether the delegation was law-
ful and whether power was exercised in accordance with the delegation. The Court
continues to discuss the relationship between NATO and the TCNs, and it is argued
that the TCN involvement – including authority concerning safety, discipline, and
accountability – was not incompatible with the effectiveness of NATO’s operational
command. At this stage the Court includes the ‘effective control’ test, in concluding
that NATO holds effective control over KFOR vis-à-vis the TCNs. This argument is
relevant because power was delegated to NATO, and not to the TCNs. Accordingly,
if the TCNs had interfered with NATO’s operational control, they could have been
held responsible for the actions on the basis of having acted outside of the scope of
the delegation. But again this raises new questions. If the Court considers that TCN
involvement amounting to effective operational control would result in attribution to
those TCNs, why does the Court not discuss this from the perspective of dual attribu-
tion? And why does the Court not say, as it did in, inter alia, the Waite and Kennedy,69
67
The Court also points to the obligatory final clause where the Security Council decides to remain ‘actively
seized of the matter’. This clause, albeit sometimes without the word ‘actively’, is included in all or practi-
cally all Security Council resolutions, and it can in my opinion not be said to carry any legal consequences.
68
Supra note 66.
69
App. No. 26083/94 Waite and Kennedy v. Germany, judgment of 18 Feb. 1999, at para. 67: ‘where States
establish international organizations … and where they attribute to these organizations certain com-
petences and accord them immunities, there may be implications as to the protection of fundamental
rights. It would be incompatible with the purpose and object of the Convention, however, if the Contract-
ing States were thereby absolved from their responsibility under the Convention in relation to the field of
activity covered by such attribution.’524 EJIL 19 (2008), 509–531
Matthews,70 and Bosphorus cases,71 that the Contracting States cannot be absolved
from their obligations under the ECHR by establishing an international organization
which carries out functions that would otherwise be carried out by the state?72
The Court is vested with the authority to ‘ensure the observance of the engagements
undertaken by the High Contracting Parties’: see Article 19 ECHR. As such, the rel-
evant subject-matter for the Court’s assessment is the conduct of Contracting States.
It therefore appears insufficient when the Court ends its discussion by concluding that
conduct is attributable to the UN. This is not what the Court is asked to decide. The
Court is asked to decide whether one or more Contracting States can be held respon-
sible, which makes it necessary to discuss whether the action can be attributable to
at least one such state. The Court’s abrupt conclusion would have been appropriate if
single attribution were the only possible solution, as the establishment of attribution
to the UN would then rule out attribution to NATO and/or TCNs. But it can be argued
that multiple attribution is not ruled out, as the citations from the commentaries on
ASR and DARIO show, in which case the Court would need to include this aspect
in the discussion. Even if it is considered that the concept of multiple attribution is
unclear and disputed, or that the UN holds such a position in international law that
multiple attribution is ruled out if a specific action is attributable to the organization,
the Court would still have to discuss the concept with a view to determining its appli-
cability. It is surprising that the Court, which in other situations has been known to
stretch its competence, the interpretation of the ECHR, and the application of general
principles under international law very far in order to ensure the effective protection
of human rights, in this case seems to interpret these same elements as narrowly as
possible in order to avoid a conclusion of admissibility.
Before we turn to the next section, it may be in order to make a disclaimer. This
article does not address the command and control structures of KFOR, and it is not
intended to draw any conclusions about the correct assessment of attribution. Assum-
ing that the ‘ultimate authority and control’ test is appropriate, the Court’s conclusion
appears to be correct. There is indeed an element of ultimate control involved, as the
Security Council has the competence to revoke the mandate or to amend it. It may
well also be the case that multiple attribution is restricted to very narrow situations, if
70
Matthews v. UK, supra note 58, at para. 32: ‘[t]he Convention does not exclude the transfers of compe-
tences to international organizations provided that Convention rights continue to be “secured”. Member
States’ responsibility therefore continues even after such a transfer.’
71
App. No. 45036/98 Bosphorus Hava Yolları Turızm ve Ticaret Anonim Şirketi v. Ireland, Grand Chamber
judgment of 30 June 2005, at para. 154: ‘[i]n … establishing the extent to which State action can be
justified by its compliance with obligations flowing from its membership of an international organization
to which it has transferred part of its sovereignty, the Court has recognized that absolving Contracting
States completely from their Convention responsibility in the areas covered by such a transfer would be
incompatible with the purpose and object of the Convention: the guarantees of the Convention could be
limited or excluded at will thereby depriving it of its peremptory character and undermining the practical
and effective nature of its safeguards … The State is considered to retain Convention liability in respect of
treaty commitments subsequent to the entry into force of the Convention.’
72
See Cerone, ‘Minding the Gap: Outlining KFOR Accountability in Post-Conflict Kosovo’, 12 EJIL (2001)
469, at 485–487, who argues that the same principle applies to KFOR.You can also read
